Bacillus subtilis LmrA is known to be a repressor that regulates the lmrAB and yxaGH operons; lmrB and yxaG encode a multidrug resistance pump and quercetin 2,3-dioxygenase, respectively. DNase I footprinting analysis revealed that LmrA and YxaF, which are paralogous to each other, bind specifically to almost the same cis sequences, LmrA/YxaF boxes, located in the promoter regions of the lmrAB operon, the yxaF gene, and the yxaGH operon for their repression and containing a consensus sequence of AWTATAtagaNYGgTCTA, where W, Y, and N stand for A or T, C or T, and any base, respectively (three-out-of-four match [in lowercase type]). Gel retardation analysis indicated that out of the eight flavonoids tested, quercetin, fisetin, and catechin are most inhibitory for LmrA to DNA binding, whereas quercetin, fisetin, tamarixetin, and galangin are most inhibitory for YxaF. Also, YxaF bound most tightly to the tandem LmrA/YxaF boxes in the yxaGH promoter region. The lacZ fusion experiments essentially supported the above-mentioned in vitro results, except that galangin did not activate the lmrAB and yxaGH promoters, probably due to its poor incorporation into cells. Thus, the LmrA/YxaF regulon presumably comprising the lmrAB operon, the yxaF gene, and the yxaGH operon is induced in response to certain flavonoids. The in vivo experiments to examine the regulation of the synthesis of the reporter β-galactosidase and quercetin 2,3-dioxgenase as well as that of multidrug resistance suggested that LmrA represses the lmrAB and yxaGH operons but that YxaF represses yxaGH more preferentially.

No abstract

DNA microarray analysis revealed that transcription of the Bacillus subtilis yetM gene encoding a putative flavin adenine dinucleotide-dependent monooxygenase was triggered by certain flavonoids during culture and was derepressed by disruption of the yetL gene in the opposite orientation situated immediately upstream of yetM, which encodes a putative MarR family transcriptional regulator. In vitro analyses, including DNase I footprinting and gel retardation analysis, indicated that YetL binds specifically to corresponding single sites in the divergent yetL and yetM promoter regions, with higher affinity to the yetM region; the former region overlaps the Shine-Dalgarno sequence of yetL, and the latter region contains a perfect 18-bp palindromic sequence (TAGTTAGGCGCCTAACTA). In vitro gel retardation and in vivo lacZ fusion analyses indicated that some flavonoids (kaempferol, apigenin, and luteolin) effectively inhibit YetL binding to the yetM cis sequence, but quercetin, galangin, and chrysin do not inhibit this binding, implying that the 4-hydroxyl group on the B-ring of the flavone structure is indispensable for this inhibition and that the coexistence of the 3-hydroxyl groups on the B-and C-rings does not allow antagonism of YetL.